Discussion Paper No. 58 of 2006 on Monetary Policy Reaction Function for Kenya

Abstract

The objective of this study was to estimate a monetary policy reaction function (MRF) for Kenya. It sought to understand whether the Central Bank of Kenya (CBK) in its reaction to macroeconomic changes and/or disturbances, reacted consistently and in a systematic way or randomly. Quarterly data from 1966, when the CBK was established, to 2003 was used in a structural vector autoregression model including a vector error correction to analyse this phenomenon. Results indicate that the reaction of monetary policy to changes in inflation were contrary to expectations during the period 1966-1990/91 when the CBK used domestic credit as the monetary policy instrument. This reflects inconsistency in the implementation of monetary policy as it contradicts policy expectations. Further, the results show that the reaction of monetary policy to changes in the output gap were insignificant, implying that monetary policy did not take into account output stabilisation. These results imply that over the entire period of study, monetary policy has not been counter-cyclical. Results from impulse response and variance decomposition analysis reveal that over the period, monetary policy behaviour was dominated by fiscal pressures and changes in net foreign assets (NFA). This points to the need for better coordination between fiscal and monetary policy as monetary policy driven by fiscal pressures may not effectively achieve its primary objectives. Results also show that monetary policy reaction to shocks or movements in the exchange rates has not been systematic. The results further indicate that monetary policy reacts more to changes in NFAs than other macroeconomic variables.